SEMRON

New paths to decentralised, heterogenous, and federated AI Training: the SPRIND Composite Learning Challenge enters Stage 2! Training large AI models remains expensive and centralized, accessible only to a few dominant players outside of Europe. To change this, SPRIND supports groundbreaking ideas for decentralized and privacy-preserving AI model development on heterogenous hardware. Out of the seven teams which entered Stage 1, the expert jury has selected five tems to advance to Stage 2. SPRIND provides up to € 530,000 in funding over the next 12 months for each team, along with mentorship and many opportunities for the teams to leverage their collective strengths.   🚀 CONGRATULATIONS to all successful teams: 👉 exalsius (Alexander Acker, Soeren Becker) 👉 SymphonyLearn (PanocularAI, Arya Mazaheri) 👉 AETHER (Tilmann Bartsch, SEMRON) 👉 CambridgeFlower (Nicholas Lane, Daniel J. Beutel, University of Cambridge, Flower Labs) 👉 Planetary Compute Platform (Phil Rohr, DeltaWave) For more information on this SPRIND Challenge, please visit our website, which is linked in the comments 👇 And a huge thank you to our wonderful jury - Johannes Otterbach, Katharine Jarmul, Wolfgang Stille, Hans Ramsl, Gerhard Wunder and Andreas Unseld and Tian Li - for their valuable expertise. Leonard Schenk Jano Costard Marcia Holst Sebastian Berns

Couldn’t have asked for a better team to work with on this. Grateful for the recognition!

Proud to share that our paper “A Blueprint for Accurate, Energy-Efficient DNN Inference via Capacitive In-Memory Computing” won the Best Industrial Paper Award at CCMCC 2025. Authored by Alexander Lowa, Boris Murmann, and myself. We demonstrate how capacitive in-memory computing enables accurate and energy-efficient DNN inference by delivering a system level framework for optimizing memcapacitive systems. A big thanks to the team and everyone who helped turn this paper into something tangible and to the organizing comitee for the great conference. #Semiconductors #ccmcc #InMemoryComputing

We have opened our first Office in the US! 🇺🇸 Expanding SEMRON’s footprint feels like a natural step in our journey. That’s why we just established our new office in Austin, Texas. The US has long been at the heart of the semiconductor industry. From pioneering research to large-scale manufacturing and strong industrial policy, it remains a place where the future of computing is shaped in decisive ways. Recent changes across the sector underline how dynamic this environment is, and for us it is the right moment to establish a local presence. Being in Austin connects us to one of the country’s most active technology hubs, with access to partners, research institutions, and the wider community driving semiconductor innovation forward. It also gives us the opportunity to grow a team in the US and engage more closely with the people and organizations that define this ecosystem. This step is also about people. Over time, we look forward to growing our team in the US and contributing to the ecosystem locally. Building AI hardware that can scale globally requires being present where progress happens. Dresden continues to be our foundation in Europe, while Austin marks the beginning of our journey in the US. 🇩🇪 🇺🇸 #Semiconductors #DeepTech #ChipDesign #AI #Austin #USA

Excited to be at [ICML] Int'l Conference on Machine Learning all week! Looking forward to connecting with like-minded researchers and discussing the latest trends in machine learning. Let’s chat if you’re around :D

Excited to be at [ICML] Int'l Conference on Machine Learning all week! Looking forward to connecting with like-minded researchers and discussing the latest trends in machine learning. Let’s chat if you’re around :D